Holmium lasers primarily find application in urology for vaporization and enucleation of hyperplastic prostate tissue (BPH) and breaking apart kidney stones, although additional applications do exist for both soft and hard tissue targets. These infrared lasers produce 0.2 joule to 6 joule pulses at 350 ms to 750 ms pulse width and 5 pps to 80 pps at 2.08 μm to 2.14 μm for average powers ranging from 8 W to 120 W.
Holmium lasers generate multimode laser energy of particularly low M2 quality. Thermally induced refractive index gradients and birefringence in holmium laser rods distort the laser output: both beam diameter and divergence drift during use and myriad modes are generated. Higher power holmium lasers employ two or more laser heads that are combined to produce the total laser output, further reducing the beam quality, and surgical lasers are subjected to jolts and bumps in hospital corridors, freight elevators, thresholds, etc., such that focusing optics are kept as robust and simple as possible. The results of this are focal spots that are large, misshapen and unstable, varying widely in parameters from manufacturer to manufacturer, throughout a single laser's lifetime and even within a single surgical session. (Nominal laser focal spot diameters (defined at 1/e2 maximum of semi-Gaussian profiles such that about 14% of the laser energy lies outside of the nominal spot diameter) for the first pulses are typically small and circular but balloon into unstable ovals of about 275 μm to about 500 μm.)
Many of the same characteristics of holmium lasers' output that make them attractive for vaporizing both hard and soft tissues add to the challenge of safely coupling fiber optics to deliver the energy; the high energy infrared pulses vaporize most materials, from polymers to metals. Optical fibers coupled to holmium lasers are routinely damaged by misalignment of the fiber core to the laser and the fibers damage the laser optics in turn; lenses are pitted or coated with organic and inorganic debris and vapor deposits, reducing performance subtly or dramatically. Subtle damage routinely goes unrecognized until accumulations result in catastrophic failure of the laser optics (blast shield, lens, mirrors, rods) or the optical fiber (at the connector or even meters away, within the patient).
Prior art designs are directed to producing fiber optic terminations that are capable of surviving significant core overfill when coupling to the laser; for example, where overfill energy is spatially filtered and typically reflected, scattered or absorbed. Some art seeks to capture at least some of the core overfill energy within the fiber core through tapered inputs (where the fiber core is larger at the input face) and others claim to reduce or eliminate coupling to the cladding, to the exclusion of the core.
These prior attempts fail to resolve high attenuation in silica-silica fiber is at 2100 nm which is highly dependent upon the mode population distribution within the fiber. Typical silica-silica fiber attenuation ranges from 1% to 3% per meter of fiber length for core modes while cladding modes are attenuated at roughly 10% meter (depending upon the secondary cladding material). Much of the energy that is lost to attenuation leaks from the fiber, into the polymer cladding and jacket. Fibers fail catastrophically where this leaked energy is of sufficient density to melt or burn the polymer layers surrounding the silica-silica fiber: a phenomenon referred to as “burn through” in the laser surgery field.
Microbending losses due to defects at the silica core to silica cladding interface are introduced during fiber preform production. Additionally, defects at the silica cladding to polymer cladding interface, stresses induced by the EFTE to fiber bonding and dimensional variations in the core are introduced when the fiber is drawn. Contributing sources to transmission losses may be within the control of the laser fiber designer, partially; for example, by selecting the best base fiber material to work with, establishing strict dimensional limits for core and claddings, and selecting among available polymer claddings. Unfortunately, insufficient cladding thickness continues to be a significant source of attenuation in holmium laser fibers.
Furthermore, cladding modes suffer greater attenuation than low order core modes and predispose a laser fiber to burn through failure. In striving to produce fiber terminations that survive spatial overfill of the fiber core, most current holmium laser fiber designs introduce new sources of cladding mode excitation. FIG. 1 illustrates two causes of cladding mode excitation in holmium laser fibers resulting from fiber termination defects. FIG. 1A depicts a fiber 5 where the fiber axis 10 is misaligned with the laser focus axis 15 such that the fiber acceptance cone α is misaligned with the laser focus cone θ and FIG. 1B depicts an angle polished fiber face 30 where the fiber face plane 35 is not orthogonal to the laser focus axis 40 such that the acceptance cone α of the fiber 45 is misaligned with the focus cone θ.
Cladding mode excitation that is due to the laser performance or damaged optics can only been mitigated by a fiber termination design, i.e. beam blooming (FIG. 2). Beam blooming is generally the result of thermal gradients within the laser, but some prior art fiber terminations amplify this problem by reflecting a portion of the laser energy back into the laser cavity, further destabilizing it or even pitting the rod face. FIG. 2A depicts a nominal holmium laser focus where the lens 50 is selected to focus a nominal output 75 of the laser rod at the focal plane 55 such that the focal spot diameter 60 is smaller than the core 65 of the fiber 70 and the focal cone angle θ is lower than the minimum acceptance cone α of the fiber 70. When holmium laser rods heat unevenly, the refractive index of the rod changes non-uniformly, producing a variable, and typically larger, diameter beam. FIG. 2B depicts the laser focus of FIG. 2A where the output 80 of the laser rod has bloomed in diameter due to thermal lensing such that it fills more of the focusing lens 85 causing the focal cone angle θ to increase, overfilling the fiber acceptance cone α and causing the focal spot diameter 90 to increase, overfilling the core 95 of the fiber.
Where the laser output blooms, the fiber meridional mode NA may be overfilled, as in FIG. 2B, but because the fiber core is larger than the nominal laser focal spot diameter, the core is not spatially overfilled. The overfilling of the fiber acceptance angle goes unnoticed in most cases because the polymer coating over the fiber's glass cladding is able to weakly guide the angular overfill, but should such fibers be subjected to bending stress, e.g. by the surgeon wrapping the fiber about his hand to gain a good grip, or by the fiber bending at the cystoscope working channel port, or just distal to the laser connection, higher order modes will be converted to cladding modes that are poorly guided, degrading the polymer cladding in a cascade of failure that typically ends catastrophically.
FIG. 3 illustrates mode conversion (mode promotion) within an exaggerated angle, tapered input fiber (neglecting refraction at the air:glass interface for simplicity in this illustration) where higher order focal modes 120 below the maximum acceptance cone angle of the fiber) (12.7°) are reflected within the taper 105 at the core:cladding interface at 130 and are raised in angle of propagation by the taper half angle of 2.5° to 12.5°. When the promoted rays encounter the taper wall a second time 135 they are again promoted by 2.5° at the core:cladding interface. The resulting angle of 15° exceeds the silica-silica numerical aperture such that, on a subsequent encounter with the taper wall 140, the rays pass through the core:cladding interface. These rays are again reflected, but by the glass:air interface of the polymer cladding free taper, and are promoted to 17.5° and finally to 20° just prior to entering the cylindrical fiber 110. In that the un-tapered fiber is coated with a low refractive index polymer, these 20° modes will be guided as cladding modes until they are lost to attenuation, exit the distal tip of the fiber, or contribute to a burn through failure.
FIG. 4 illustrates a method of compensating for mode conversion of higher angle excited modes where the same angle higher order mode as depicted in FIG. 3 150 is refracted at 160 by a negative curvature lens 155 such that the refracted mode never encounters the taper wall 175, but instead reflects for the first time within the cylindrical fiber 170 at 165. Using such a concave lens input, tapered input fibers may perform as well as, or better than, many straight input fibers, yet these types of terminations can excite and convert cladding modes under more stressful conditions such as beam blooming or scatter in damaged optics.
Other fiber termination strategies, e.g. FIG. 5, may also inadvertently launch cladding modes. U.S. Pat. No. 7,090,411 (Brown) discloses a glass ferrule 235 surrounding a polymer denuded fiber 230 with unpolished (saw cut) glass faces 220 & 245 acting as diffusers as well as internal multifaceted reflectors and reduced diameter input fibers. Such scattering elements, as exemplified by 220 and 245, scatter laser focal rays 210 with the bulk of the overfill energy being redirected toward polymer clad 250 and ETFE buffered 265 segments of the distal fiber such that very high order scattered modes may couple to the fiber core:cladding within the polymer-free segment proximal to 215 and become guided as cladding modes within the polymer clad fiber at 250. Employing tapered fibers in the reverse of FIG. 3 will convert higher order modes to lower orders only when the taper axial alignment is assured and taper angles are lower than the highest order modes excited within the fiber core.
Accordingly, improvements in fiber termination technologies are desirable.